Use cases
Hardship assessment and forbearance
When a borrower says they're struggling, gather the context, run the back-office review, and resolve against your forbearance policies.
Tags
AI Agent
Inbound
Outbound
Back office
Channel
Chat
Voice
Category
Collections
Industry
Lending
What it does
Borrowers in financial difficulty are some of the most consequential conversations a lender has. Get the questions right and you keep the customer, stay inside Consumer Duty, and find a workout that holds. Miss the signals and you end up with a complaint, a charge-off, or worse.
The Gradient Labs AI agent picks up the hardship signal the moment it appears, on a collections call, an inbound query, or a missed payment trigger. It asks the right questions to understand the borrower's situation, triggers a back-office hardship review, and assesses the case against your forbearance policies. The borrower hears back through whatever channel suits them, with full history carried forward.

How it works
On a call, the agent listens for hardship language ("I lost my job," "I'm going through a separation," "I can't afford the full payment") and switches into a hardship sub-procedure. It asks the income, expenditure, and household questions your team would ask, captures the answers, and flags any vulnerability indicators (mental health, bereavement, financial abuse) that route the case to a specialist.
The back-office agent then reviews the case against your forbearance policies: payment holiday, reduced payments, term extension, or refer-for-review. It checks affordability, applies the eligibility rules your team set, and produces a recommendation with the supporting evidence and the rule it applied. A human approves the outcome where your policy requires it.
When the borrower calls back to check on the decision, the inbound agent already has the case loaded and walks them through the outcome. Every interaction is logged with the conversation, the evidence, and the decision rationale, ready for Consumer Duty and audit review.

See it in action
Explore other use cases